Automated model generation for integrated building and HVAC performance simulation

Abstract

With growing requirements and changing technologies, making the right choices in building design is more challenging than ever. Building performance simulation (BPS) has the potential to support the planning of more energy-effcient buildings by modeling their thermal behaviour in its complexity before their actual construction or modification. This involves taking into account heating, ventilation and air-conditioning (HVAC) systems and their interaction with the building and its occupants. Still, obtaining a detailed evaluation of the performance of building and HVAC systems in early planning stages is made diffcult by cumbersome model preparation. While building geometry from building information models (BIM) can be translated for use in building performance simulation, such model transformations are more problematic when it comes to HVAC data. Moreover, detailed descriptions of HVAC systems whose performance is to be assessed are generally not available during those early stages in which simulation could be most useful. The approach developed in this thesis aims at avoiding this diffculty by creating models of plausible HVAC systems corresponding to available building models. Parametric procedures for the automated creation of such models are presented, and combined into a method allowing integrated building and HVAC simulation models to be derived from available building models. A software system prototype realizing the method is developed. Different efforts aiming at the validation of the system are presented. System results are subjected to comparative testing. The effects of various model simplifications are investigated. It is shown how simulation calibration can be carried out with the system when measurements are available. Finally, the application of the system to building and HVAC design use cases is illustrated. By making detailed assessments of the integrated performance of building and HVAC concepts more affordable, the method can fill a gap in the use of simulation for decision support in the conceptual design phase.